US20110145437A1 - Operation of a content distribution network - Google Patents
Operation of a content distribution network Download PDFInfo
- Publication number
- US20110145437A1 US20110145437A1 US13/058,442 US200913058442A US2011145437A1 US 20110145437 A1 US20110145437 A1 US 20110145437A1 US 200913058442 A US200913058442 A US 200913058442A US 2011145437 A1 US2011145437 A1 US 2011145437A1
- Authority
- US
- United States
- Prior art keywords
- content
- server
- access server
- access
- servers
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/58—Association of routers
- H04L45/586—Association of routers of virtual routers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L61/00—Network arrangements, protocols or services for addressing or naming
- H04L61/35—Network arrangements, protocols or services for addressing or naming involving non-standard use of addresses for implementing network functionalities, e.g. coding subscription information within the address or functional addressing, i.e. assigning an address to a function
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/2866—Architectures; Arrangements
- H04L67/289—Intermediate processing functionally located close to the data consumer application, e.g. in same machine, in same home or in same sub-network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/563—Data redirection of data network streams
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L61/00—Network arrangements, protocols or services for addressing or naming
- H04L61/09—Mapping addresses
- H04L61/25—Mapping addresses of the same type
- H04L61/2503—Translation of Internet protocol [IP] addresses
- H04L61/2546—Arrangements for avoiding unnecessary translation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/56—Provisioning of proxy services
- H04L67/568—Storing data temporarily at an intermediate stage, e.g. caching
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/40—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass for recovering from a failure of a protocol instance or entity, e.g. service redundancy protocols, protocol state redundancy or protocol service redirection
Definitions
- This invention relates to the operation of content distribution networks.
- One measure that content providers can take is to make the content available from a number of separate content caches instead of a single central server.
- the individual caches are periodically updated from a master database or server, and may or may not store the same data, (e.g. there is the capability to tailor the content to local conditions, for example different linguistic preferences of the populations of the areas served by each cache).
- An individual user requiring a data download is directed to one of the caches from which the data can be downloaded.
- the user would request the URL (universal resource location) code of the content provider, e.g. www.contentprovider.co.qq (where qq is a country code).
- This URL is processed by the domain name system (DNS) server co-operating with the user's browser system to identify a network-compatible internet address, typically a 32 or 128 bit number (e.g. IP address) which identify the network location of the target computer.
- DNS domain name system
- the domain name system is a set of hierarchical servers. Each DNS server stores a subset of all the correspondences between URLs and IP addresses.
- a DNS server does not have a record for a particular URL, the required information is sought from the authoritative DNS server for the domain associated with that URL, which either returns the required IP address or itself refers the request to another DNS server in its hierarchy, and so on.
- the required correspondence once retrieved from a higher level server, is then recorded in all the lower level DNS servers requesting it, thereby allowing more efficient retrieval on subsequent requests.
- the URLs for download streaming services and other sites that may be cached are widely advertised and therefore have to be valid for a wide area, generally worldwide.
- the structure of the internet, and the way in which DNS servers operate, generally require a common IP address to be generated in response to a given URL. If a number of separate content caches are to be provided and accessed efficiently, a way needs to be found to allow the user to be given access to the appropriate one. This could be done by allowing the content provider to redirect users to the appropriate cache, but this would require the content provider to be able to identify where in the network topology the user is located. This is not readily apparent from the user identity (IP address)—the situation may also change dynamically depending on mobility of the user or changes in the configuration of the network. In any case, efficient routing of traffic takes place in the network, and primarily benefits the network operator. It is not necessarily within the content provider's area of expertise. For these reasons a network-based solution would be desirable.
- One existing method of directing internet traffic is to implement a routing protocol such as the border gateway protocol (BGP) to maintain a routing policy which maintains a table of IP network addresses, whereby data is routed to the “nearest” or “best” of these destinations as viewed by the routing topology.
- BGP border gateway protocol
- a system known as “anycast” (by analogy with unicast, broadcast, and multicast) is also sometimes used to enable geographically distributed nodes to share a single IP address.
- each destination address identifies a set of some or (in broadcast) all receivers in the network as endpoints, but unlike either of these other systems only one of the set of endpoints (the “nearest” or “best”) is selected to receive a transmission at any one time.
- This arrangement requires the routing protocol to maintain the routing list to determine which of the set of receivers is currently the “nearest” or “best” for each user or access point. Changes in network topology, mobility of network users, or other factors require such frequent changes to such a routing list.
- the architecture of a modern telecommunications network consists of a number of relatively self-contained subdomains, usually serving different geographical areas although these need not be defined rigidly.
- a typical example discussed in European patent application EP1331788 provides each subdomain with its own unique IP address. It is known for each such subdomain to have more than one content distribution server associated with it, each such server in a given subdomain having identical content. In such arrangements a load-balancing function is used to determine which of the servers is to service each request for content.
- the load balancer can provide access using the same network address to any of the servers associated with a given subdomain using the same network address. Examples of such load balancers are discussed in United States Patent application US2006/0133371 and International Patent application WO06/072114.
- the subdomains are interconnected both to allow communication between them and to provide robustness, so that in the event of technical difficulties in the equipment serving one subdomain, it can access processing power from and/or obtain connectivity through a neighbouring one.
- the present invention makes use of this architecture to provide improved access to a content distribution system.
- Each subdomain can be associated with a specified content cache. Note that this need not be a one-to-one correspondence, as one cache may serve more than one subdomain and one subdomain may be served by more than one cache. Because of the architecture of the network, the cache associated with the same subdomain as the requesting user will be the topologically closest, and therefore the most efficient for it to access.
- a content distribution system comprising a plurality of access servers arranged to route traffic to and from end users on a network, each of said access servers having means for transmitting data packets to an associated content server or group of content servers, wherein all of the content servers in said content distribution system are accessible using the same internet protocol address, and each access server is configured to route any data packets it receives, if addressed to the said internet protocol address, to its respective content server.or group of content servers
- the invention provides a method of operating a content distribution system arranged to route traffic between end users and a plurality of content servers all accessible using the same internet protocol (IP) address, by way of a plurality of access servers, wherein each access server is configured to identify the internet protocol address common to the content servers as being associated with a respective content server or group of content servers, and wherein on receipt of a data packet from one of said end users addressed to the said internet protocol address, the access server routes that request to its associated content server or group of content servers.
- IP internet protocol
- This architecture differs from the prior art systems in that a common IP address is used not only for all the content servers within a given subdomain, as is known for load balancing, all the separate subdomains also share a network address. This allows the same IP address to be used for access to the service, regardless of which subdomain the user makes contact with. This has advantages both in publicising the network address (for example through a hyperlink in a related webpage) and for consistency of access for mobile users who can use the same network address to access the local content server wherever they may connect to the network.
- the content server is a provider of content for downloading, or streaming, to the end user in response to a data packet or packets incorporating a request for that content.
- access is preferably by means of a load balancer or other means for distributing data requests among the group.
- each access server is associated with a particular content server or group of such servers, each access server can be configured in the same way to route data packets to its associated content server even though there are multiple content servers with the same IP address. It will recognise the common IP address as relating to its associated content server and route content requests accordingly. All the content servers have the same IP address, but this is not a problem as each access server needs only to be able to reach one content server, although an access server may be able to reach more than one content server in which case it will determine which one to route content requests to based on the routing list that it maintains. The user will get the same content whichever access server it makes contact with, because all the content servers have access to the same content (either cached locally or via another content server).
- each access server can determine, based on the routing list maintained by the access server, to which content server requests should be routed.
- the network is divided into a plurality of network subdomains, each of which is associated with a corresponding access server, and its associated content server. Delivery of the content required by the user is therefore confined to one subdomain.
- one content server may serve more than one network division, which may itself be accessed by more, than one access server.
- Each access server will only ever identify the IP address with the content server with which it is associated.
- each access server directs data packets to the content server having the nearest IP-routeable point to the source of the request for content, which is topologically the closest to it and its clients.
- the system is preferably configured such that in the event of a content delivery criterion failure condition being detected, all traffic passing through one access server can be redirected to another content server by way of another access server in said content distribution system. This is preferably effected by the provision of routing apparatus arranged to implement said redirection when said failure condition is detected.
- this is achieved by rerouting (e.g. by means of a “tunnel”) the request from the initial access server to a second access server elsewhere, the second access server being associated with a different content server.
- the second access server will recognise the IP address as relating to its own associated content server and direct the request thereto.
- the content server will nevertheless route the downloaded material correctly to the requesting user.
- FIG. 1 is a schematic and simplified representation of a network incorporating the invention
- FIG. 1 a shows an alternative embodiment
- FIG. 2 is a flow diagram illustrating the operation of the system in normal use
- FIG. 3 is a flow diagram illustrating the operation of the system in the event of failure of one of the content caches.
- FIG. 1 a shows a first, sub-optimal embodiment.
- a plurality of access servers 10 , 20 , 30 each have an associated content cache (content server) 13 , 23 .
- Two or more access servers 20 , 30 may be associated with the same content cache 23 .
- the caches 13 , 23 all have the same IP address.
- a request from a content consumer 11 , 12 , 21 , 31 will be routed to one of the access servers 10 , 20 , 30 .
- the access server is not pre-determined: routing may depend on load-balancing and other network effects.
- the access server 10 , 20 , 30 routes requests for content to its associated cache.
- each access server 10 , 20 , 30 Since each access server 10 , 20 , 30 has its own designated cache, 13 , 23 the duplication of IP addresses will not result in any misrouting because each access server 10 , 20 , 30 will have no record or knowledge of any cache other than the one with which it is associated.
- This arrangement is not the preferred embodiment, because the requesting device 21 may access any of several access servers 10 , 20 , 30 and delivery of content will be from whichever cache 13 , 23 is associated therewith. This may not be the optimal delivery route.
- a preferred embodiment is shown in FIG. 1 .
- FIG. 1 depicts a simplified partitioned network having several subdomains 1 , 2 , 3 , each having a respective access server 10 , 20 , 30 and each serving a plurality of end users 11 , 12 ; 21 , 22 ; 31 , 32 .
- Each end user connects to the access server which is associated with the geographic network sub-domain in which it is located.
- the network sub-domain may be a layer-1, layer-2 or layer-3 network.
- Each access server 10 , 20 , 30 is connected to a respective content cache 13 , 23 or a set of such caches, which stores content and downloads it to end users on request.
- a content cache 13 may be co-located with an access server 10 , or it may be located elsewhere in the network.
- a content cache 13 may be dedicated to a single access server 10 , or alternatively one content cache 23 may be shared between two or more access servers 20 , 30 .
- an alternative routing 19 , 29 may be available between an access server 10 and a content cache 23 other than its associated cache 13 , for use in exceptional circumstances.
- each such content cache may in practice consist of multiple devices 130 , 131 , 132 etc all connected to the access server(s) through a load-balancing device 14 which apportions the demand appropriately between the devices 130 , 131 , 132 .
- the access servers 10 , 20 , 30 are also connected to other networks 4 such as Internet Service Providers, or the Internet itself, or corporate intranets). These external networks 4 may have direct connectivity to each network sub-domain 1 , 2 , 3 , as shown, or they may have aggregated connectivity to the network sub-domains through one or more common interface mediums.
- networks 4 such as Internet Service Providers, or the Internet itself, or corporate intranets.
- All the content caches 13 , 23 are given the same IP address, referred to herein as “W.X.Y.Z”.
- the cache 13 comprises several subcaches 130 , 131 , 132 fronted by a load balancer 14 .
- the load balancer 14 that has the common IP address. This means that the configuration of the access servers 10 , 20 , 30 to handle content caching/distribution/streaming can be the same regardless of its location in the network.
- FIG. 2 The operation of the system in normal use is depicted in FIG. 2 .
- a user requiring the streaming services is told the URL (universal resource location) of a piece of content, for example through a link in a web page, an embedded URL object in a web page, or some other means) and enters the URL code e.g. “www.contentsupplier.co.qq/content” into the browser facility of his terminal 11 .
- the user's browser 11 requests (step 201 ), from its local DNS server 5 , the IP address associated with the DNS name of that URL (i.e. the section of the URL code before the “/”). Instead of a browser, access may be by way of a dedicated client, or some other means.
- the DNS system 5 returns the IP address “W.X.Y.Z” associated with that URL (step 202 ). Note that a unique IP address is associated with the URL, as is conventional, and it is that IP address which is returned to the user 11 , regardless of where in the network the user is located. However, this IP address is applicable to all the individual content caches 13 , 23 .
- the end user's configured DNS content server 5 returns an alternative DNS name shared by all the content caches/streamers (e.g. cdn.company.co.qq).
- the end user application 11 requests the IP address associated with this shared DNS name.
- the DNS server 5 is configured with a list of IP addresses associated with the DNS name in the request. One of these is returned to the End User, the one selected being chosen based on any number of mechanisms and factors such as for load balancing purposes.
- the list of IP addresses could consist of only a single IP address, e.g. the IP address shared by all the content caches/streamers (i.e. W.X.Y.Z).
- the user's browser 11 Having received the IP address (step 202 ) the user's browser 11 then transmits a request ( 203 ), addressed to the IP address “W.X.Y.Z”.
- This request is routed to the nearest access server 10 , which uses standard routing/policy mechanisms to direct end user traffic addressed to that IP address to the content cache 13 associated with that access server 10 (step 204 ). It can do this without having to take into account the address or geographic location of the end user 11 , because each access server 10 operates exclusively in a geographic sub-domain, and so the subdomain 1 to which any end user 11 , 12 is connected is necessarily identified as the one associated with the access server 10 .
- Each access server 10 ( 20 , 30 ) recognises the IP address W.X.Y.Z as relating uniquely to its respective associated content cache 13 , ( 23 , 23 ) which, because of the architecture of the network, is necessarily the closest to any user terminal 11 , ( 21 , 31 ) transmitting a data request to that access server.
- each access server can be configured in the same way, and no address translation or mapping is required to route requests to the correct local content server.
- the content cache 13 connected to the End User's access server 10 may send a redirect to the browser 11 to cause it to retrieve the content from a different content server 23 , which may or may not be co-located with the original content/cache.
- the content server to which the End User is redirected may have an address S.T.U.V, different from the shared content caching IP address (W.X.Y.Z).
- an access server 10 may be connected to more than one content cache 13 , 23 , either directly or via another device such as another access server 20 which in turn is connected to a content cache 23 .
- This allows the system to remain operable in the event of failure of a first content cache 13 or of the connection between the access node 10 and the first content cache 13 . Operation of the system in such an event is depicted in FIG. 3 .
- the access server 10 If the access server 10 is unable to connect to the associated content server 13 , it is configured to forward traffic to the second content cache 23 . It is not possible for the first access server 10 to redirect requests by simple routing because all the content caches 23 , 43 have the same IP address W.X.Y.Z, and so the forwarded request would simply be returned to the local content cache 13 . Instead, in the event of a failure ( 305 ) to transmit the request to the local cache 13 , the request is forwarded by “tunnelling” the request to the second access server 20 . This could be done by generating a packet (step 306 ), for transmission to the second access server 20 , including the original request and the cache address W.X.Y.Z, to the network IP address of the second access server 20 .
- the forwarding address applied by the first access server 10 is that of the second access server 20 .
- This request is therefore transmitted to the second access server 20 (step 307 ).
- the second access server 20 extracts the data request and identifies the address of the request as “W.X.Y.Z”.
- the second access server 20 recognises this address as being that of its own associated content server 23 , rather than the content server 13 associated with the first access server 10 , and forwards the request accordingly (step 308 ).
- the content can then be returned to the address of the requesting user (step 309 ).
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Information Transfer Between Computers (AREA)
Abstract
Description
- This invention relates to the operation of content distribution networks.
- The distribution of video and other streamed content to users on demand using the internet as a distribution medium has become a major industry. Efficient distribution systems are becoming necessary in order to manage the sheer volume of data to be carried over the network.
- One measure that content providers can take is to make the content available from a number of separate content caches instead of a single central server. The individual caches are periodically updated from a master database or server, and may or may not store the same data, (e.g. there is the capability to tailor the content to local conditions, for example different linguistic preferences of the populations of the areas served by each cache). An individual user requiring a data download is directed to one of the caches from which the data can be downloaded.
- However, such a system is only efficient if each user request can be directed to the most appropriate cache, normally the closest available one. In existing systems this would require a translation to be made between the network address of the end user (in the initial request message) and the address of the local cache.
- In a typical system, the user would request the URL (universal resource location) code of the content provider, e.g. www.contentprovider.co.qq (where qq is a country code). This URL is processed by the domain name system (DNS) server co-operating with the user's browser system to identify a network-compatible internet address, typically a 32 or 128 bit number (e.g. IP address) which identify the network location of the target computer. The domain name system is a set of hierarchical servers. Each DNS server stores a subset of all the correspondences between URLs and IP addresses. If a DNS server does not have a record for a particular URL, the required information is sought from the authoritative DNS server for the domain associated with that URL, which either returns the required IP address or itself refers the request to another DNS server in its hierarchy, and so on. In most arrangements, the required correspondence, once retrieved from a higher level server, is then recorded in all the lower level DNS servers requesting it, thereby allowing more efficient retrieval on subsequent requests.
- The URLs for download streaming services and other sites that may be cached are widely advertised and therefore have to be valid for a wide area, generally worldwide. The structure of the internet, and the way in which DNS servers operate, generally require a common IP address to be generated in response to a given URL. If a number of separate content caches are to be provided and accessed efficiently, a way needs to be found to allow the user to be given access to the appropriate one. This could be done by allowing the content provider to redirect users to the appropriate cache, but this would require the content provider to be able to identify where in the network topology the user is located. This is not readily apparent from the user identity (IP address)—the situation may also change dynamically depending on mobility of the user or changes in the configuration of the network. In any case, efficient routing of traffic takes place in the network, and primarily benefits the network operator. It is not necessarily within the content provider's area of expertise. For these reasons a network-based solution would be desirable.
- One existing method of directing internet traffic is to implement a routing protocol such as the border gateway protocol (BGP) to maintain a routing policy which maintains a table of IP network addresses, whereby data is routed to the “nearest” or “best” of these destinations as viewed by the routing topology. A system known as “anycast” (by analogy with unicast, broadcast, and multicast) is also sometimes used to enable geographically distributed nodes to share a single IP address. Like broadcast (one-to-all) and multicast (one-to-many), in “anycast” each destination address identifies a set of some or (in broadcast) all receivers in the network as endpoints, but unlike either of these other systems only one of the set of endpoints (the “nearest” or “best”) is selected to receive a transmission at any one time. This arrangement requires the routing protocol to maintain the routing list to determine which of the set of receivers is currently the “nearest” or “best” for each user or access point. Changes in network topology, mobility of network users, or other factors require such frequent changes to such a routing list.
- The architecture of a modern telecommunications network consists of a number of relatively self-contained subdomains, usually serving different geographical areas although these need not be defined rigidly. A typical example discussed in European patent application EP1331788 provides each subdomain with its own unique IP address. It is known for each such subdomain to have more than one content distribution server associated with it, each such server in a given subdomain having identical content. In such arrangements a load-balancing function is used to determine which of the servers is to service each request for content. Thus the load balancer can provide access using the same network address to any of the servers associated with a given subdomain using the same network address. Examples of such load balancers are discussed in United States Patent application US2006/0133371 and International Patent application WO06/072114.
- The subdomains are interconnected both to allow communication between them and to provide robustness, so that in the event of technical difficulties in the equipment serving one subdomain, it can access processing power from and/or obtain connectivity through a neighbouring one. The present invention makes use of this architecture to provide improved access to a content distribution system. Each subdomain can be associated with a specified content cache. Note that this need not be a one-to-one correspondence, as one cache may serve more than one subdomain and one subdomain may be served by more than one cache. Because of the architecture of the network, the cache associated with the same subdomain as the requesting user will be the topologically closest, and therefore the most efficient for it to access.
- According to a first aspect of the invention, there is provided a content distribution system comprising a plurality of access servers arranged to route traffic to and from end users on a network, each of said access servers having means for transmitting data packets to an associated content server or group of content servers, wherein all of the content servers in said content distribution system are accessible using the same internet protocol address, and each access server is configured to route any data packets it receives, if addressed to the said internet protocol address, to its respective content server.or group of content servers
- According to a second aspect, the invention provides a method of operating a content distribution system arranged to route traffic between end users and a plurality of content servers all accessible using the same internet protocol (IP) address, by way of a plurality of access servers, wherein each access server is configured to identify the internet protocol address common to the content servers as being associated with a respective content server or group of content servers, and wherein on receipt of a data packet from one of said end users addressed to the said internet protocol address, the access server routes that request to its associated content server or group of content servers.
- This architecture differs from the prior art systems in that a common IP address is used not only for all the content servers within a given subdomain, as is known for load balancing, all the separate subdomains also share a network address. This allows the same IP address to be used for access to the service, regardless of which subdomain the user makes contact with. This has advantages both in publicising the network address (for example through a hyperlink in a related webpage) and for consistency of access for mobile users who can use the same network address to access the local content server wherever they may connect to the network.
- In a preferred embodiment, the content server is a provider of content for downloading, or streaming, to the end user in response to a data packet or packets incorporating a request for that content. Where a group of content servers are associated with one access server, access is preferably by means of a load balancer or other means for distributing data requests among the group.
- Because each access server is associated with a particular content server or group of such servers, each access server can be configured in the same way to route data packets to its associated content server even though there are multiple content servers with the same IP address. It will recognise the common IP address as relating to its associated content server and route content requests accordingly. All the content servers have the same IP address, but this is not a problem as each access server needs only to be able to reach one content server, although an access server may be able to reach more than one content server in which case it will determine which one to route content requests to based on the routing list that it maintains. The user will get the same content whichever access server it makes contact with, because all the content servers have access to the same content (either cached locally or via another content server).
- The duplication of IP addresses is possible because each access server can determine, based on the routing list maintained by the access server, to which content server requests should be routed.
- In a preferred embodiment, the network is divided into a plurality of network subdomains, each of which is associated with a corresponding access server, and its associated content server. Delivery of the content required by the user is therefore confined to one subdomain. Note however that one content server may serve more than one network division, which may itself be accessed by more, than one access server. Each access server will only ever identify the IP address with the content server with which it is associated. Preferably each access server directs data packets to the content server having the nearest IP-routeable point to the source of the request for content, which is topologically the closest to it and its clients.
- To provide robustness to the system the system is preferably configured such that in the event of a content delivery criterion failure condition being detected, all traffic passing through one access server can be redirected to another content server by way of another access server in said content distribution system. This is preferably effected by the provision of routing apparatus arranged to implement said redirection when said failure condition is detected.
- Preferably this is achieved by rerouting (e.g. by means of a “tunnel”) the request from the initial access server to a second access server elsewhere, the second access server being associated with a different content server. As the content servers both have the same IP address, the second access server will recognise the IP address as relating to its own associated content server and direct the request thereto. However, as the requesting user has a unique IP address, the content server will nevertheless route the downloaded material correctly to the requesting user.
- An embodiment of the invention will now be described by way of example and with reference to the Figures, in which
-
FIG. 1 is a schematic and simplified representation of a network incorporating the invention -
FIG. 1 a shows an alternative embodiment -
FIG. 2 is a flow diagram illustrating the operation of the system in normal use -
FIG. 3 is a flow diagram illustrating the operation of the system in the event of failure of one of the content caches. -
FIG. 1 a shows a first, sub-optimal embodiment. In this embodiment a plurality ofaccess servers more access servers same content cache 23. Thecaches content consumer access servers access server access server access server - This arrangement is not the preferred embodiment, because the requesting
device 21 may access any ofseveral access servers cache FIG. 1 . -
FIG. 1 depicts a simplified partitioned network havingseveral subdomains respective access server end users - Each
access server respective content cache content cache 13 may be co-located with anaccess server 10, or it may be located elsewhere in the network. Acontent cache 13 may be dedicated to asingle access server 10, or alternatively onecontent cache 23 may be shared between two ormore access servers alternative routing access server 10 and acontent cache 23 other than its associatedcache 13, for use in exceptional circumstances. - As shown for
content cache 13, each such content cache may in practice consist ofmultiple devices device 14 which apportions the demand appropriately between thedevices - The
access servers other networks 4 such as Internet Service Providers, or the Internet itself, or corporate intranets). Theseexternal networks 4 may have direct connectivity to eachnetwork sub-domain - All the
content caches FIG. 1 , thecache 13 comprisesseveral subcaches load balancer 14. In such a case it is theload balancer 14 that has the common IP address. This means that the configuration of theaccess servers - The operation of the system in normal use is depicted in
FIG. 2 . A user requiring the streaming services is told the URL (universal resource location) of a piece of content, for example through a link in a web page, an embedded URL object in a web page, or some other means) and enters the URL code e.g. “www.contentsupplier.co.qq/content” into the browser facility of histerminal 11. The user'sbrowser 11 requests (step 201), from itslocal DNS server 5, the IP address associated with the DNS name of that URL (i.e. the section of the URL code before the “/”). Instead of a browser, access may be by way of a dedicated client, or some other means. - The
DNS system 5 returns the IP address “W.X.Y.Z” associated with that URL (step 202). Note that a unique IP address is associated with the URL, as is conventional, and it is that IP address which is returned to theuser 11, regardless of where in the network the user is located. However, this IP address is applicable to all theindividual content caches - In an alternative arrangement, the end user's configured
DNS content server 5 returns an alternative DNS name shared by all the content caches/streamers (e.g. cdn.company.co.qq). Theend user application 11 then requests the IP address associated with this shared DNS name. TheDNS server 5 is configured with a list of IP addresses associated with the DNS name in the request. One of these is returned to the End User, the one selected being chosen based on any number of mechanisms and factors such as for load balancing purposes. The list of IP addresses could consist of only a single IP address, e.g. the IP address shared by all the content caches/streamers (i.e. W.X.Y.Z). - Having received the IP address (step 202) the user's
browser 11 then transmits a request (203), addressed to the IP address “W.X.Y.Z”. This request is routed to thenearest access server 10, which uses standard routing/policy mechanisms to direct end user traffic addressed to that IP address to thecontent cache 13 associated with that access server 10 (step 204). It can do this without having to take into account the address or geographic location of theend user 11, because eachaccess server 10 operates exclusively in a geographic sub-domain, and so thesubdomain 1 to which anyend user access server 10. Each access server 10 (20, 30) recognises the IP address W.X.Y.Z as relating uniquely to its respective associatedcontent cache 13, (23, 23) which, because of the architecture of the network, is necessarily the closest to anyuser terminal 11, (21, 31) transmitting a data request to that access server. Thus each access server can be configured in the same way, and no address translation or mapping is required to route requests to the correct local content server. - If the
content cache 13 connected to the End User'saccess server 10 cannot itself deliver the required content to theEnd User 11, either because it does not hold the data or because it is overloaded with similar requests, it may send a redirect to thebrowser 11 to cause it to retrieve the content from adifferent content server 23, which may or may not be co-located with the original content/cache. The content server to which the End User is redirected may have an address S.T.U.V, different from the shared content caching IP address (W.X.Y.Z). - As shown in
FIG. 1 , anaccess server 10 may be connected to more than onecontent cache access server 20 which in turn is connected to acontent cache 23. This allows the system to remain operable in the event of failure of afirst content cache 13 or of the connection between theaccess node 10 and thefirst content cache 13. Operation of the system in such an event is depicted inFIG. 3 . - If the
access server 10 is unable to connect to the associatedcontent server 13, it is configured to forward traffic to thesecond content cache 23. It is not possible for thefirst access server 10 to redirect requests by simple routing because all thecontent caches 23, 43 have the same IP address W.X.Y.Z, and so the forwarded request would simply be returned to thelocal content cache 13. Instead, in the event of a failure (305) to transmit the request to thelocal cache 13, the request is forwarded by “tunnelling” the request to thesecond access server 20. This could be done by generating a packet (step 306), for transmission to thesecond access server 20, including the original request and the cache address W.X.Y.Z, to the network IP address of thesecond access server 20. Thus the forwarding address applied by thefirst access server 10 is that of thesecond access server 20. This request is therefore transmitted to the second access server 20 (step 307). On receiving the forwarded data request, thesecond access server 20 extracts the data request and identifies the address of the request as “W.X.Y.Z”. Thesecond access server 20 recognises this address as being that of its own associatedcontent server 23, rather than thecontent server 13 associated with thefirst access server 10, and forwards the request accordingly (step 308). The content can then be returned to the address of the requesting user (step 309).
Claims (14)
Applications Claiming Priority (7)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP08252819.1 | 2008-08-26 | ||
EP08252815.9 | 2008-08-26 | ||
EP08252815A EP2159994A1 (en) | 2008-08-26 | 2008-08-26 | Operation of a content distribution network |
EP08252819 | 2008-08-26 | ||
EP08252815 | 2008-08-26 | ||
EP08252819A EP2159991A1 (en) | 2008-08-26 | 2008-08-26 | Accessing a content distribution network |
PCT/GB2009/001952 WO2010023424A1 (en) | 2008-08-26 | 2009-08-07 | Operation of a content distribution network |
Publications (2)
Publication Number | Publication Date |
---|---|
US20110145437A1 true US20110145437A1 (en) | 2011-06-16 |
US9203921B2 US9203921B2 (en) | 2015-12-01 |
Family
ID=41607132
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/058,442 Active 2030-07-19 US9203921B2 (en) | 2008-08-26 | 2009-08-07 | Operation of a content distribution network |
Country Status (3)
Country | Link |
---|---|
US (1) | US9203921B2 (en) |
EP (1) | EP2319229B1 (en) |
WO (1) | WO2010023424A1 (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8700945B1 (en) * | 2011-09-23 | 2014-04-15 | Amazon Technologies, Inc. | Content delivery failover |
US20190045014A1 (en) * | 2016-01-29 | 2019-02-07 | Tectonic Interactive Limited | System and method for managing communication sessions between clients and a server |
US10574788B2 (en) * | 2016-08-23 | 2020-02-25 | Ebay Inc. | System for data transfer based on associated transfer paths |
CN113678126A (en) * | 2019-03-26 | 2021-11-19 | 谷歌有限责任公司 | Separate authorization for content access and content delivery using multiple encrypted digital signatures |
US11902383B2 (en) * | 2018-11-26 | 2024-02-13 | Neoatomic LLC | Adaptive media content delivery network |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8898287B2 (en) * | 2010-02-24 | 2014-11-25 | Salesforce.Com, Inc. | System, method and computer program product for monitoring data activity utilizing a shared data store |
US8745122B2 (en) * | 2011-06-14 | 2014-06-03 | At&T Intellectual Property I, L.P. | System and method for providing an adjunct device in a content delivery network |
CN104320487B (en) * | 2014-11-11 | 2018-03-20 | 网宿科技股份有限公司 | The HTTP scheduling system and method for content distributing network |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20020004846A1 (en) * | 2000-04-28 | 2002-01-10 | Garcia-Luna-Aceves J. J. | System and method for using network layer uniform resource locator routing to locate the closest server carrying specific content |
US20020091760A1 (en) * | 2001-01-10 | 2002-07-11 | John Rozen | Distributed selection of a content server |
US20030079027A1 (en) * | 2001-10-18 | 2003-04-24 | Michael Slocombe | Content request routing and load balancing for content distribution networks |
US20030182410A1 (en) * | 2002-03-20 | 2003-09-25 | Sapna Balan | Method and apparatus for determination of optimum path routing |
US20030233423A1 (en) * | 2002-04-09 | 2003-12-18 | Dilley John A. | Method and system for tiered distribution in a content delivery network |
US20050010653A1 (en) * | 1999-09-03 | 2005-01-13 | Fastforward Networks, Inc. | Content distribution system for operation over an internetwork including content peering arrangements |
US20060064478A1 (en) * | 2004-05-03 | 2006-03-23 | Level 3 Communications, Inc. | Geo-locating load balancing |
US20060133371A1 (en) * | 2004-12-20 | 2006-06-22 | Fujitsu Limited | Communication data relay system and method |
US20070006270A1 (en) * | 2005-06-29 | 2007-01-04 | Nortel Networks Limited | Timely recovery for media on demand streaming |
US20070160058A1 (en) * | 2006-01-09 | 2007-07-12 | Weiqiang Zhou | Method and system for implementing backup based on session border controllers |
US20070168548A1 (en) * | 2006-01-19 | 2007-07-19 | International Business Machines Corporation | Method and system for performing multi-cluster application-specific routing |
US20070271385A1 (en) * | 2002-03-08 | 2007-11-22 | Akamai Technologies, Inc. | Managing web tier session state objects in a content delivery network (CDN) |
US20080082642A1 (en) * | 2006-10-02 | 2008-04-03 | Futurewei Technologies, Inc. | Context Transfer and Common IP Address for DHCP Proxy Solution in WiMAX |
US20100034079A1 (en) * | 2007-02-06 | 2010-02-11 | France Telecom | Service management in a network |
US20100036954A1 (en) * | 2008-08-06 | 2010-02-11 | Edgecast Networks, Inc. | Global load balancing on a content delivery network |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003223378A (en) * | 2002-01-29 | 2003-08-08 | Fujitsu Ltd | Contents delivery network service method and system |
-
2009
- 2009-08-07 US US13/058,442 patent/US9203921B2/en active Active
- 2009-08-07 EP EP09784898.0A patent/EP2319229B1/en active Active
- 2009-08-07 WO PCT/GB2009/001952 patent/WO2010023424A1/en active Application Filing
Patent Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050010653A1 (en) * | 1999-09-03 | 2005-01-13 | Fastforward Networks, Inc. | Content distribution system for operation over an internetwork including content peering arrangements |
US20020004846A1 (en) * | 2000-04-28 | 2002-01-10 | Garcia-Luna-Aceves J. J. | System and method for using network layer uniform resource locator routing to locate the closest server carrying specific content |
US20020091760A1 (en) * | 2001-01-10 | 2002-07-11 | John Rozen | Distributed selection of a content server |
US20030079027A1 (en) * | 2001-10-18 | 2003-04-24 | Michael Slocombe | Content request routing and load balancing for content distribution networks |
US20080235400A1 (en) * | 2001-10-18 | 2008-09-25 | Level 3 Communications, Llc | Content request routing and load balancing for content distribution networks |
US20070271385A1 (en) * | 2002-03-08 | 2007-11-22 | Akamai Technologies, Inc. | Managing web tier session state objects in a content delivery network (CDN) |
US20030182410A1 (en) * | 2002-03-20 | 2003-09-25 | Sapna Balan | Method and apparatus for determination of optimum path routing |
US20030233423A1 (en) * | 2002-04-09 | 2003-12-18 | Dilley John A. | Method and system for tiered distribution in a content delivery network |
US20070055764A1 (en) * | 2002-04-09 | 2007-03-08 | Dilley John A | Method and system for tiered distribution in a content delivery network |
US20060064478A1 (en) * | 2004-05-03 | 2006-03-23 | Level 3 Communications, Inc. | Geo-locating load balancing |
US20060133371A1 (en) * | 2004-12-20 | 2006-06-22 | Fujitsu Limited | Communication data relay system and method |
US20070006270A1 (en) * | 2005-06-29 | 2007-01-04 | Nortel Networks Limited | Timely recovery for media on demand streaming |
US20070160058A1 (en) * | 2006-01-09 | 2007-07-12 | Weiqiang Zhou | Method and system for implementing backup based on session border controllers |
US20070168548A1 (en) * | 2006-01-19 | 2007-07-19 | International Business Machines Corporation | Method and system for performing multi-cluster application-specific routing |
US20080082642A1 (en) * | 2006-10-02 | 2008-04-03 | Futurewei Technologies, Inc. | Context Transfer and Common IP Address for DHCP Proxy Solution in WiMAX |
US20100034079A1 (en) * | 2007-02-06 | 2010-02-11 | France Telecom | Service management in a network |
US20100036954A1 (en) * | 2008-08-06 | 2010-02-11 | Edgecast Networks, Inc. | Global load balancing on a content delivery network |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8700945B1 (en) * | 2011-09-23 | 2014-04-15 | Amazon Technologies, Inc. | Content delivery failover |
US9379849B2 (en) | 2011-09-23 | 2016-06-28 | Amazon Technologies, Inc. | Content delivery failover |
US20190045014A1 (en) * | 2016-01-29 | 2019-02-07 | Tectonic Interactive Limited | System and method for managing communication sessions between clients and a server |
US10819801B2 (en) * | 2016-01-29 | 2020-10-27 | Tectonic Interactive Limited | System and method for managing communication sessions between clients and a server |
US10574788B2 (en) * | 2016-08-23 | 2020-02-25 | Ebay Inc. | System for data transfer based on associated transfer paths |
US11316953B2 (en) | 2016-08-23 | 2022-04-26 | Ebay Inc. | System for data transfer based on associated transfer paths |
US11902383B2 (en) * | 2018-11-26 | 2024-02-13 | Neoatomic LLC | Adaptive media content delivery network |
CN113678126A (en) * | 2019-03-26 | 2021-11-19 | 谷歌有限责任公司 | Separate authorization for content access and content delivery using multiple encrypted digital signatures |
Also Published As
Publication number | Publication date |
---|---|
EP2319229A1 (en) | 2011-05-11 |
EP2319229B1 (en) | 2018-09-12 |
WO2010023424A1 (en) | 2010-03-04 |
US9203921B2 (en) | 2015-12-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP2319221B1 (en) | Content distribution network | |
US11336614B2 (en) | Content node network address selection for content delivery | |
US6415323B1 (en) | Proximity-based redirection system for robust and scalable service-node location in an internetwork | |
US9203921B2 (en) | Operation of a content distribution network | |
EP1865684B1 (en) | A content distribution method over an internetwork including content peering arrangement | |
US8214524B2 (en) | System and method for selecting an optimal authoritative name server | |
US8806008B2 (en) | HTML delivery from edge-of-network servers in a content delivery network (CDN) | |
US20030079027A1 (en) | Content request routing and load balancing for content distribution networks | |
EP2159994A1 (en) | Operation of a content distribution network | |
KR20130039652A (en) | Method and apparatus for load balancing of content centric network | |
EP2159991A1 (en) | Accessing a content distribution network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: BRITISH TELECOMMUNICATIONS PUBLIC LIMITED COMPANY, Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:NIVEN-JENKINS, BENJAMIN PAUL;BURKE, PAUL WILLIAM;SIGNING DATES FROM 20090915 TO 20090917;REEL/FRAME:025789/0756 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1551); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 4 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 8 |